Downloads · 30 days
0
GA8Q10
ChetanNaharIITM/GA8Q10
GA8Q10 is a machine learning model from ChetanNaharIITM. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
This repository documents the carbon accounting for an assigned GPU training run.
Access
Public
Updated Aug 28, 2026
Repo size
—
Likes
0
Public
Repo makeup
Click a slice to open those files.
Other1.5 KB · 51%
At a glance
- Access
- Public
- Created
- Aug 28, 2026
- Updated
- Aug 28, 2026
- SHA
- 821ee20b
Model card
From the Hugging Face model README
Green AI Carbon Accounting Audit
Overview
This repository documents the carbon accounting for an assigned GPU training run.
The training run used an NVIDIA L40S GPU configuration for fine-tuning.
Training Information
| Parameter | Value |
|---|---|
| GPU | NVIDIA L40S |
| Number of GPUs | 7 |
| GPU hours | 305.3 |
| GPU TDP | 350 W |
| PUE | 1.41 |
| Region | asia-south1 |
| Grid carbon intensity | 650 gCO2eq/kWh |
| Training type | fine-tuning |
Energy Calculation
The total energy consumption was calculated using:
energy_kWh = (GPU TDP × number of GPUs × GPU hours × PUE) / 1000
Substituting the values:
energy_kWh = (350 × 7 × 305.3 × 1.41) / 1000
Therefore:
Total Energy = 1054.659 kWh
Carbon Emissions Calculation
Carbon emissions were calculated using:
co2_kg = (energy_kWh × regional carbon intensity) / 1000
Substituting the values:
co2_kg = (1054.659 × 650) / 1000
Therefore:
Total Carbon Emissions = 685.528 kg CO2eq
Green AI Notes
The carbon estimate accounts for GPU power consumption, the number of GPUs, GPU runtime, datacenter Power Usage Effectiveness (PUE), and the regional electricity grid carbon-intensity value supplied in the assignment.
The calculated emissions are recorded in the Hugging Face Model Card YAML frontmatter above.
- Created
- Aug 28, 2026
- Updated
- Aug 28, 2026